Modeling, History Matching, Forecasting and Analysis of Shale Reservoirs Performance Using Artificial Intelligence

spineunkemptΤεχνίτη Νοημοσύνη και Ρομποτική

17 Ιουλ 2012 (πριν από 5 χρόνια και 1 μήνα)

1.127 εμφανίσεις


SPE 143875
Modeling, History Matching, Forecasting and Analysis of Shale Reservoirs
Performance Using Artificial Intelligence
Shahab D. Mohaghegh, Intelligent Solutions, Inc. & West Virginia University, Ognjen Grujic, Seed Zargari, and
Masoud Kalantari, West Virginia University
Copyright 2011, Society of Petroleum Engineers

This paper was prepared for presentation at the SPE Digital Energy Conference and Exhibition held in The Woodlands, Texas, USA, 19–21 April 2011.

This paper was selected for presentation by an SPE program committee following review of information contained in an abstract submitted by the author(s). Contents of the paper have not been
reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect any position of the Society of Petroleum Engineers, its
officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written consent of the Society of Petroleum Engineers is prohibited. Permission to
reproduce in print is restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of SPE copyright.


Abstract
Producing hydrocarbon from Shale plays has attracted much attention in the recent years. Advances in horizontal
drilling and multi-stage hydraulic fracturing have made shale reservoirs a focal point for many operators. Our
understanding of the complexity of the flow mechanism in the natural fracture and its coupling with the matrix
and the induced fracture, impact of geomechanical parameters and optimum design of hydraulic fractures has not
necessarily kept up with our interest in these prolific and hydrocarbon rich formations.
In this paper we discuss using a new and completely different approach to modeling, history matching,
forecasting and analyzing oil and gas production in shale reservoirs. In this new approach instead of imposing our
understanding of the flow mechanism and the production process on the reservoir model, we allow the production
history, well log, and hydraulic fracturing data to force their will on our model and determine its behavior. In
other words, by carefully listening to the data from wells and the reservoir we developed a data driven model and
history match the production process from Shale reservoirs. The history matched model is used to forecast future
production from the field and to assist in planning field development strategies. We use the last several months of
production history as blind data to validate the model that is developed.
This is a unique and innovative use of pattern recognition capabilities of artificial intelligence and data mining as
a workflow to build a full field reservoir model for forecasting and analysis of oil and gas production from shale
formations. Examples of three case studies in Lower Huron and New Albany shale formations (gas producing)
and Bakken Shale (oil producing) is presented in this paper.

Introduction
This article reviews a new reservoir simulation and modeling technology called Top-Down, Intelligent Reservoir
Modeling (Top-Down Modeling - TDM - or short) as it is applied to shale formations with examples presented for
New Albany, Lower Huron and Bakken Shales. The natural fractures in the shale contribute significantly to the
production as the main conduit for reservoir permeability. Recent revival of interest in production from shale
formations can be attributed to multi-stage hydraulic fractures. It is a known fact that success of these hydraulic
fracturing procedures is directly related to their ability to reach and intersect the existing natural fractures in the
shale formation. Mapping of the natural fractures in the shale formations have proven to be an elusive task. Even
with most advanced logging technologies one can only detect the intersection of the natural fractures with the
wellbore while the extent of these fracture beyond the wellbore and how they are distributed throughout the
reservoir (between wells) remains the subject of research.
Top-Down Modeling tries to model the impact of the hydraulic fractures and natural fractures on the production
from wells rather than modeling the discrete fracture networks. While developing stochastic realizations of
natural fractures and their intersection with the induced hydraulic fracturing are being studied using stochastic and
2 Modeling of Shale Reservoirs Performance Using Artificial Intelligence SPE 143875
numerical reservoir modeling, TDM fills the existing gap for a predictive model that can be built using minimum
amount of assumptions about the nature of the reservoir and our understanding of its complexity. TDM starts with
a solid assumption that whatever the nature of the natural fracture distribution and it s interaction with the induced
hydraulic fractures may be, their impact is bound to show itself in the amount of the hydrocarbon that each well is
able to produce. These signatures can be used in order to build reservoir models, match the production history and
build a predictive model that can help us making reservoir management decisions.
Top-Down Modeling technology is an elegant integration of traditional reservoir engineering methods with
pattern recognition capabilities of artificial intelligence and data mining. Advantages of this new modeling
technology include its flexible data requirement, short development time and ease of development and analysis.
Its shortcoming is that it can only be applies to brown fields where reasonable amount of data from the field in
accessible. The data requirements for the Top-Down Modeling necessitate a field with about 35 to 40 wells and
about 5 years of production history. As number of wells increases, the amount of required production history may
be reduced.
Traditional reservoir simulation and modeling is a bottom-up approach. It starts with building a geological model
of the reservoir followed by adding engineering fluid flow principles (Diffusivity equation, Darcy’s law, Fick’s
law of diffusion …) to arrive at a dynamic reservoir model. The dynamic reservoir model is calibrated using the
production history of multiple wells and the history matched model is used to strategize field development in
order to improve recovery.
Top-Down Modeling approaches the reservoir simulation and modeling from an opposite angle by attempting to
build a realization of the reservoir starting with well production behavior (history). The production history is
augmented with core, log, well test and seismic data (upon availability of each) in order to increase the accuracy
and fine tune the Top-Down Model. The model is then calibrated (history matched) using the most recent wells as
blind dataset. Although not intended as a substitute for the traditional reservoir simulation of large, complex
fields, this novel approach can be used as an alternative (at a fraction of the cost and time) to traditional numerical
reservoir simulation in cases where performing traditional modeling is cost (and man-power) prohibitive,
specifically for shale formations. In cases where a conventional model of a reservoir already exists, Top-Down
Modeling should be considered a complement to, rather than a competition for the traditional technique. It
provides an independent look at the data coming from the reservoir/wells for optimum development strategy and
recovery enhancement.
Top-Down Modeling provides a unique perspective of the field and the reservoir using actual measurements. It
provides qualitatively accurate reservoir characteristics maps that can play a key role in making important and
strategic field development decisions.

Top-Down, Intelligent Reservoir Modeling for Shale Formations
Traditional reservoir simulation is the industry standard for reservoir management. It is used in all phases of field
development in the oil and gas industry and is now being used on some but not all of the shale formations. The
routine of simulation studies calls for integration of static and dynamic measurements into the reservoir model. It
is a bottom-up approach that starts with building a geological (geo-cellular or static) model of the reservoir. Using
modeling and geo-statistical manipulation of the data the geo-cellular model is populated with the best available
petrophysical and geophysical information. Engineering fluid flow principles are added and solved numerically to
arrive at a dynamic reservoir model. The dynamic reservoir model is calibrated using the production history of
multiple wells in a process called history matching and the final history matched model is used to strategize the
field development in order to improve recovery. Characteristics of the traditional reservoir simulation and
modeling include:
1. It takes a significant investment (time and money) to develop a geological (geo-cellular, static) model to
serve as the foundation of the reservoir simulation model.
2. Development and history matching of a reservoir simulation model is not a trivial process and requires
modelers and geoscientists with significant amount of experience.
3. It is an expensive and time consuming endeavor.
SPE 143875 Mohaghegh, Grujic, Zargari, & Kalantari 3
4. A prolific asset is required in order to justify a significant initial investment that is required for a reservoir
simulation model.
Top-Down Intelligent Reservoir
Modeling can serve as an alternative or
a complement to traditional reservoir
simulation and modeling. It starts with
well-known reservoir engineering
techniques such as Decline Curve
Analysis, Type Curve Matching,
History Matching using single well
numerical reservoir simulation,
Volumetric Reserve Estimation and
calculation of Recovery Factors for all
the wells (individually) in the field.
Using statistical techniques multiple
Production Indicators (3, 6, and 9
months cumulative production as well
as 1, 3, 5, and 10 year cumulative
production) are calculated. The
reservoir engineering analyses along
with the statistical data form the basis for a comprehensive spatio-temporal database. This database represents an
extensive set of snap shots of fluid flow in the shale formation. It is expected that all the characteristics that
governs the complexity of fluid flow in the naturally fractured reservoir to be captured in this extensive spatio-
temporal database. This large volume of data is processed using the state-of-the-art in artificial intelligence and
data mining (neural modeling, genetic optimization and fuzzy pattern recognition) in order to generate a complete
and cohesive model of the entire reservoir. This is accomplished by using a set of discrete modeling techniques to
generate production related predictive models of well behavior, followed by intelligent models that integrate the
discrete models into a cohesive model of the reservoir as a whole, using a continuous fuzzy pattern recognition
algorithms.
The Top-Down, Intelligent Reservoir Model is calibrated using the most recent set of wells that have been drilled
in the field. The calibrated model is then used for field development strategies to improve and enhance
hydrocarbon recovery. In the following sections some of the results that have been achieved from application of
TDM to three shale formations are briefly presented.
Top-Down Models are used in reservoir management workflows using the flowchart that is shown in Figure 1.
Upon completion of the spatio-temporal database, which proves to be one of the most important steps in
development of a Top-Down Model
(TDM), the process of training and
history matching of the TDM is
performed simultaneously. It must be
noted that a rigorous blind history
matching is required in this step of the
process to ensure the robustness of the
Top-Down Model. Using the design tool
that is part of the TDM process, field
development strategies are planed and
then using the history matched model (in
predictive mode) the plans are tested to
see if they fulfill the objectives of
reservoir management. This process is
repeated, iteratively (by planning new
wells to be drilled and predicting their
performance), until the reservoir
Figure
1
.

Reservoir management

workflow using Top
-
Down
Modeling
.


Figure
2
.

Shale distribution in Kentucky
.


4 Modeling of Shale Reservoirs Performance Using Artificial Intelligence SPE 143875
management objectives are met. Once the objective is accomplished, the plan is forwarded to operation for
implementation. The Top-Down Model, like any other reservoir model, needs to be updated regularly, as shown
in the flow chart in Figure 1. It is noteworthy to mention that most of the work presented in this paper has been
performed on publicly available data. Only for Lower Huron shale some completion data was acquired from one
of the operators in the region.

A
PPLICATION OF
T
OP
-D
OWN
M
ODELING TO
L
OWER
H
URON
S
HALE

While the details of the Top-Down modeling application to Lower Huron Shale can be found in the SPE paper
(Grujic 2010) some new information on this study are presented here. Thickness of the shale formations in
Kentucky are shown in Figure 2, identifying the deeper and thicker shale formations. Depth, formation thickness
and porosity distribution of the portion of the field that is the subject of Top-Down Modeling is shown in Figure
3.

Figure 3. Formation Depth, Pay Thickness and Porosity distribution used in the TDM.
Following the flow chart that was presented in Figure 1, the TDM for the Lower Huron is trained and history
matched. During the TDM training and history matching part of the production history (usually the tail-end of the
production) is removed from the model building process and is used as blind test in order to check the validity of
the reservoir model. The quality of the TDM is usually judged based on its capability to predict the part of the
production history that has not been used during the reservoir model training.
Figure 4 shows the strategy that was
incorporated during the Top-Down
Model training, history matching and
blind history matching for the Lower
Huron Shale. Production history was
available for this field from 1982 to
2008. The Top-Down Model was
trained and history matched with data
from 1982 to 2004 and production
history from 2005 to 2008 was left out
to be used as validation of the model in
the form of blind history matching.
Figure 5 shows the result of training and
history matching of the Top-Down
Model for Lower Huron shale when
applied to the production history of the
entire field (this study included a
portion of a field with 75 wells). Top-Down Model is built (trained and history matched) on a well by well basis
and in order to generate the plot in Figure 5, both production history and TDM results had to be combined for all
Figure
4
.

Strategy used during the training and history
matching of the
Top-Down Model for the Lower Huron Shale.

SPE 143875 Mohaghegh, Grujic, Zargari, & Kalantari 5
the wells in the study. In this figure
result of TDM is compared with the
actual production history from the
field in monthly production rate
versus time and cumulative field
production versus time. To
demonstrate the results of TDM on
single wells two examples are
presented in Figure 6.
This figure shows the results of TDM
model training and history matching
(blind portion of history matching is
shown in different color) for two
wells, namely well KF1184 and
KF1638. These figures demonstrate
the predictive capability of TDM in
Lower Huron shale.
As demonstrated in the flow chart of
Figure 1, TDM includes a design
module. The objective of the design
module is to assist in performing
reservoir management tasks such as
identifying which portion of the
reservoir has been depleted. By identifying reservoir depletion as a function of time (which is a reflection of
pressure draw down in the field) and by cross referencing that with the original hydrocarbon in place, an
indication of remaining reserves in the field (as a function of time and well placement) will emerge.


Figure 6. Results of the Top-Down Modeling (monthly rate and cumulative production) as applied to the production history
from wells KH1184 and KF 1638.


Figure
5
.

Results of the TDM (monthly rate and cumulative production) as
applied to the production from the entire field.

6 Modeling of Shale Reservoirs Performance Using Artificial Intelligence SPE 143875
TDM design tool uses Fuzzy Pattern Recognition in order to identify the portions of the shale formation that has
contributed the most to the production during the first three month, 3, 5 and 10 years as shown in Figure 7. Details
of this Fuzzy Pattern Recognition process has been covered in several previously published papers (Gomez 2009
– Kalantari 2009 – Kalantari 2010 – Mata 2007 – Mohaghegh 2009). In this figure the reservoir is delineated into
several RRQIs (Relative Reservoir Quality Index) shown in different colors. The portion of the reservoir that is
shown with the darkest color represents RRQI of 1. This is the portion of the reservoir that has made the largest
contribution to production followed by RRQI 2, 3, 4. The colors of other RRQIs gradually get lighter until the
region for RRQI 5 become almost white
The contribution of the
delineated RRQIs to
production is calculated
taking into account the
number of wells that are
included in each of the
RRQIs. Furthermore, these
regions refer to depletion
in the shale formation
since locations that have
the highest amount of
production are, relatively
speaking, the most
depleted parts of the
reservoir. Figure 7 shows
that in this field the central
part is the most depleted
portion of the reservoir
with more depletion shown
in the north and south parts
of the field. Furthermore, it
shows that as time
progresses the most
depleted central portion of
the reservoir expands
toward east and west.
The design tool in the Top-
Down Modeling that is
powered by Fuzzy Pattern
Recognition technology is
used to support reservoir
management decisions
such as identifying infill
locations in the field.
For instance, in order to
calibrate the location of the
reservoir quality separator
lines in Figure 7 the latest
drilled wells in the field
(wells drilled in 2008) are
removed from the analysis
and the reservoir
delineations is performed
using wells prior to 2008.
Then the production
Figure
7
.

Results of the TDM’s Fuzzy Pattern Recognition showing reservoir depletion
as a function of time.
Figure 8. Testing the validity of identified RRQIs in Lower Huron Shale.
SPE 143875 Mohaghegh, Grujic, Zargari, & Kalantari 7
indicator for the wells that are drilled in 2008 are
compared with the RRQI that they are located in to
find out if the pattern recognition analysis (reservoir
delineation into different RRQIs) is valid.
This exercise is performed on the first year
cumulative production of wells completed in the
Lower Huron Shale. Figure 8 shows that the first year
cumulative production of wells drilled in RRQI(2)
should be between 27.3 and 39.9 MMSCF and the
first year cumulative production of wells drilled in
RRQI(3) should be between 18.7 and 27.3 MMSCF .
The averaged first year cumulative productions of
wells drilled in RRQI (2) in 2008 were 33.9 MMSCF
while the averaged first year cumulative production of
wells drilled in RRQI (3) in 2008 were 22.0 MMSCF, both within the predicted range.

A
PPLICATION OF
T
OP
-D
OWN
M
ODELING TO
B
AKKEN
S
HALE


While the details of Top-Down modeling application to Bakken Shale can be found in a recently published SPE
paper (Zargari 2010) some new information about this study is presented here. Similar modeling and analyses
were performed for both Upper and Middle Bakken. A combination of both of these studies is presented here.
Figure 9 shows the portion of the field with wells that have been completed in Upper and Middle Bakken.
Voronoi polygons have been generated for the wells in this field.


Figure 11. Distribution of Pay Thickness and Porosity in the Upper Bakken Shale.
Figure
9
.

Voronoi polygons identified for

the wells in
Upper Bakken and Middle Bakken.
Figure
10
.

Strategy used during the training and history
matching of the Top-Down Model for the Bakken Shale.
8 Modeling of Shale Reservoirs Performance Using Artificial Intelligence SPE 143875
Figure 10 shows the strategy that was incorporated during the Top-Down Model training, history matching and
blind history matching for the Bakken Shale. Figures 11 and 12 show the distribution of some reservoir
characteristics in Upper (Figure 11) and Middle Bakken (Figure 12). In Top-Down Modeling a high-level static
model of the reservoir is developed based on well logs and all other available reservoir characteristics. Since
TDM is an AI&DM-based reservoir simulation and modeling technology it does not require a static model in the
form that is common and customary for the numerical reservoir simulation models.


Figure 12. Distribution of Pay Thickness and Deep Resistivity in the Middle Bakken.

The static model that is developed during the Top-Down Modeling process uses only the available (and preferably
measured) data. The objective is to refrain from interpretations, as much as possible. The TDM static model
represents the reservoir characteristic indications (well logs, results of core analysis and well tests, seismic
attributes) that is associated with each well and relates them with similar reservoir characteristic indications from
the offset wells. By performing this for all the wells the reservoir characteristic indications of each portion of the
reservoir is sampled multiple times, once as the main well and several times as offset to the neighboring wells.


Figure 13. Training and history matching of Top-Down Modeling in Upper Bakken Shale.

SPE 143875 Mohaghegh, Grujic, Zargari, & Kalantari 9
Figures 13 and 14 show the performance of the Top-Down Model after training and history matching for both
Upper Bakken Shale and Middle Bakken. In each of these figures four examples are shown. From these figures it
can be concluded that TDM has captured the essence of fluid flow in naturally fractured shale reservoirs and can
model (in predictive mode) the performance of wells in such formations.
One of the capabilities of Top-Down Model is its ability to perform fast track analysis. As part of such analyses
TDM is capable of developing type curves for each of the wells in order to quantify (in predictive mode and for
new wells) the uncertainties associated with parameters that are used as input to the model. Such parameters can
be reservoir characteristics or operational constraints that are imposed on the well during production. If
parameters involved in the hydraulic fracturing such as number of stages or amount of proppant injected are part
of the input parameters of the TDM, they can also be used during such analyses.
Figure 15 shows an example of such analysis that can be performed routinely once a Top-Down Model is trained
and history matched for a shale formation. In this figure production rate is plotted against time for a given well
while the formation thickness (on the left) and Porosity (on the right) are changed. TDM shows the expected
changes in production behavior in each of these wells as formation thickness and porosity are modified.


Figure 14. Training and history matching of Top-Down Modeling in Middle Bakken.


Figure 15. Results of the Top-Down Modeling (monthly rate and cumulative production) as applied to the production
history.


10 Modeling of Shale Reservoirs Performance Using Artificial Intelligence SPE 143875
Similar to the analysis that was
presented for the Lower Huron
shale, the TDM design tool can be
used in order to analyze the
depletion in the shale reservoir and
identify the remaining reserves.
Figure 16 shows the contribution
of different part of Upper Bakken
shale to production as a function of
time. It can be seen that during the
first 3 years of production
contribution to production is
concentrated on the south-eastern
part of the field while as times
goes on the south-western and
western part of the field starts to
contribute more and more until it
becomes the dominant contributor
to the production by the end of the
tenth year of production. As these
contributions to the production
(depletion) are cross referenced
with the original oil in place, a
qualitative picture of remaining
reserves in the field starts to
emerge. Figure 17 shows the remaining reserves in this part of the Upper Bakken Shale as of January of 2010.
Maps such as the one shown in Figure 17 can play an important role in reservoir management decisions that are
made in Bakken Shale.

A
PPLICATION OF
T
OP
-D
OWN
M
ODELING TO
N
EW
A
LBANY
S
HALE

Detail of Top-Down Modeling application to New Albany Shale can be found in a recently published SPE paper
(Kalantari 2009). Figure 18 shows the location of the New Albany Shale and the portion of the formation that was
used in the Top-Down Modeling
along with well locations, the
Cartesian and the Voronoi grid
that was used to identify the
Estimated Ultimate Drainage
Area (EUDA) for each well. The
limitations imposed on this study
included the extent of the
publicly available data (some
production history along with
well logs for a subset of wells).
Once the TDM static model was
constructed, the initial gas in
place was calculated and
mapped. As mentioned in the
prior sections, this is an
important first step in
development process of Top-
Down Models. This figure
(Figure 18) also shows the
permeability and the initial gas in
Figure
17
.
Remaining Reserves as of January 2010 in the Upper Bakken Shale.
Figure
16
.

Depletion in the Upper Bakken Shale as a function of time using the
Fuzzy Pattern Recognition of Top-Down Modeling.
SPE 143875 Mohaghegh, Grujic, Zargari, & Kalantari 11
place distribution determined using the type curve matching and volumetric calculations in the Top-Down
modeling workflow. The permeability distribution in this part of New Albany Shale was calculated using a history
matching process that involved dynamic modeling of the production from some of the wells using a stochastic
discrete fracture network model. Details of this procedure have been covered in the original SPE paper (Kalantari
2009).
Figure 19 shows the Remaining Reserves as a function of time from 2006 to 2040 (if no new wells are drilled in
this part of the field). Estimation of the Remaining Reserves is one of the last steps in the Top-Down Modeling
work flow. Remaining Reserves along with other outputs from the Top-Down Modeling provides means for
identifying the optimum locations for infill (new) wells. Using the trained and history matched model that is
developed during one of the earlier stages of Top-Down Modeling, production from infill wells can be predicted
(estimated) and the Remaining Reserves under new circumstances may be calculated and mapped similar to those
in Figure 19.

Figure 18. Applying Top-Down Modeling to New Albany Shale (NAS). Steps involved in preparing the model.


Figure 19. Assessing Remaining Reserves as a function of time Using Fuzzy Pattern Recognition, NAS.

Conclusions
Share of shale formations to overall hydrocarbon production in the United States and in the world is increasing
rapidly. As the interest in production from shale increases so does the interest in managing shale reservoir and
consequently building predictive reservoir simulation model for shale formations. Modeling shale reservoirs is a
complex process. Contribution of concentration gradient dependent diffusion along with fluid flow through
12 Modeling of Shale Reservoirs Performance Using Artificial Intelligence SPE 143875
discrete natural fracture networks become even more complex with multi-stage hydraulic fractures that are used
for completing wells in shale reservoirs. All these factors make reservoir modeling of shale formation particularly
difficult and challenging.
Probably the most challenging part of the shale reservoir modeling is our quest for accurate representation of the
natural fracture network and the intersection of the induced fractures with these networks. In this paper we
introduce a novel approach for modeling hydrocarbon producing shale reservoirs by concentrating on production
history and any and all available reservoir characteristics measurements and operational constraints. In this
approach we follow the philosophy of doing the best with the available information and trying to stay away from
assumptions about our understanding of the details of what actually has happened in the formation. In this
modeling approach, instead of starting from first principle physics, we let the actual physics impose itself on the
final model through data. This data driven modeling technology concentrates on measured data rather than
assumptions.
In this paper we demonstrated the application of Top-Down, Intelligent Reservoir Modeling (TDM) to gas
producing Lower Huron Shale and New Albany Shale and oil producing Bakken Shale. We showed the results of
trained and history matched models in matching actual production from the field including blind history matches.
We reviewed that application of the design tool that is offered by TDM in identification of infill location as well
as depletion in the reservoir and remaining reserves. Top-Down, Intelligent Reservoir Modeling (TDM) is a
technically viable alternative to numerical reservoir simulation that can be performed at a fraction of the cost and
man power in order to help engineers and geoscientists learn more about shale formations and try to manage such
reservoirs.

References
• Gomez, Y., Khazaeni, Y., Mohaghegh, S.D., and Gaskari, R. 2009 "Top-Down Intelligent Reservoir
Modeling (TDIRM)". SPE 124204, Proceedings, 2009 SPE Annual Conference & Exhibition. New
Orleans, Louisiana.
• Grujic, O., Mohaghegh, S.D. and Bromhal, G. 2010 “Fast Track Reservoir Modeling of Shale Formations
in the Appalachian Basin. Application to Lower Huron Shale in Eastern Kentucky”. SPE 139101.
Proceedings 2010 SPE Eastern Regional Conference & Exhibition. 12-14 October 2010. Morgantown,
West Virginia.
• Kalantari-Dahaghi, A.M., Mohaghegh, S.D. 2009 "Top-Down Intelligent Reservoir Modeling of New
Albany Shale". SPE 125859, Proceedings, 2009 SPE Eastern Regional Conference & Exhibition.
Charleston, West Virginia.
• Kalantari-Dahaghi, A.M., Mohaghegh, S.D. and Khazaeni, Y. 2010 "New Insight into Integrated
Reservoir Management using Top-Down, Intelligent Reservoir Modeling Technique; Application to a
Giant and Complex Oil Field in the Middle East". SPE 132621, Proceedings, 2010 SPE Western Regional
Conference & Exhibition. 27-29 May 2010. Anaheim, California.
• Mata, D., Gaskari, R., Mohaghegh, S.D., 2007 "Field-Wide Reservoir Characterization Based on a New
Technique of Production Data Analysis". SPE 111205, Proceedings, 2007 SPE Eastern Regional
Conference & Exhibition. 17-19 October 2007. Lexington, Kentucky.
• Mohaghegh, S.D. 2009 "Top-Down Intelligent Reservoir Modeling (TDIRM); A New Approach In
Reservoir Modeling By Integrating Classic Reservoir Engineering With Artificial Intelligence & Data
Mining Techniques". AAPG 2009 Annual Convention and Exhibition. June 7-10, 2009. Denver,
Colorado.
• Zargari, S., Mohaghegh, S.D. and Bromhal, G. 2010 “Field Development Strategies for Bakken Shale
Formation”. SPE 139032. Proceedings 2010 SPE Eastern Regional Conference & Exhibition. 12-14
October 2010. Morgantown, West Virginia.